Recommendation Questions Revisited

Recently there was a commercial from a plastic bag company. The commercial was about how great their plastic bags were and how widely they’re used across the country. At the end of the commercial, the narrator says: “… and according to our survey, 73% of all [our bag] users would recommend us to friends and family.” At first glance, that number may seem high. 73% is nearly three quarters of all of its users, and it’s always good to know that the people using your product or service are likely to recommend it to others. But after thinking about that number for a while, it may not be very high at all, and while the details of this particular study are not known, there are reasons that this number may highlight problems with interpreting survey results.

Perhaps the most noticeable issue is that it’s only referring to customers that use the product regularly. It can be assumed that those that use the product are going to be recommending it, otherwise they would have gone to another competitor’s product a while ago. Thinking of it in those terms, 73% seems fairly low.

It’s not often that someone that uses a product all the time would not recommend it to others. Customer satisfaction surveys are valuable tools, but one has to imagine that, on average, when a recurring customer has to pick between two choices where one is favorable to the company and one is unfavorable, most people are likely going to choose the favorable option. It seems (although this may be wrong) that those that would have answered this question would be more likely to choose that they would recommend a product that they may otherwise not care much about.

Once again, this makes it seem as though 73% is fairly low. There is no way to know exactly if these numbers are truly “low.” It is possible that the study itself shows that 73% is a fairly high number based on the sample, how the research was conducted, etc.

But after thinking about it, this number certainly doesn’t seem very high. It stands to reason that there are likely companies that include this question in their surveys only to completely misread the data and come to the wrong conclusions, and it goes to highlight how important it is that not only is a survey run correctly, but also that every result (both favorable and unfavorable) is being thought about critically.